2017ApJ...842..128K

The ratio of the [C II] 158 µm emission line over the total infrared emission (TIR) is often used as a proxy for the photoelectric (PE) heating efficiency (ε_PE) of the far-ultraviolet (FUV) photons absorbed by dust in the interstellar medium. In the nearby galaxy M31, we measure a strong radial variation of [C II]/TIR that we rule out as being due to an intrinsic variation in ε_PE. [C II]/TIR fails as a proxy for ε_PE, because the TIR measures all dust heating, not just the contribution from FUV photons capable of ejecting electrons from dust grains. Using extensive multi-wavelength coverage from the FUV to far-infrared, we infer the attenuated FUV emission (UV_att), and the total attenuated flux (TOT_att). We find [C II]/TIR to be strongly correlated with UV_att/TOT_att, indicating that, in M31 at least, one of the dominant drivers for [C II]/TIR variation is the relative hardness of the absorbed stellar radiation field. We define ε_PE^UV, [C II]/UV_att which should be more closely related to the actual PE efficiency, which we find to be essentially constant (1.85±0.8 % ) in all explored fields in M31. This suggests that part of the observed variation of [C II]/TIR in other galaxies is likely due to a change in the relative hardness of the absorbed stellar radiation field, caused by a combination of variations in the stellar population, dust opacity, and galaxy metallicity, though PE efficiency may also vary across a wider range of environments.